HAWAII GEOTHERMAL PROJECT University of Hawaii Holmes Hall 240 2540 Dole Street - Honolulu, Hawaii 96822 WELL TEST AND RESERVOIR ENGINEERING PROGRESS REPORT FOR JANUARY 1977 SUPPORT FOR PHASE III PROVIDED BY: Energy Research and Development Administration State of Hawaii Paul C. Yuen, Engineering Bill H. Chen, Hilo College Deane H. Kihara, Mechanical Engineering Department Patrick K. Takahashi, Civil Engineering Department University of Hawaii Honolulu, Hawaii 96822 .J
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
HAWAII GEOTHERMAL PROJECT
University of HawaiiHolmes Hall 240 2540 Dole Street
- Honolulu, Hawaii 96822
WELL TEST AND RESERVOIR ENGINEERING
PROGRESS REPORT FOR JANUARY 1977
SUPPORT FOR PHASE III PROVIDED BY:
Energy Research and Development AdministrationState of Hawaii
Paul C. Yuen, EngineeringBill H. Chen, Hilo College
Deane H. Kihara, Mechanical Engineering DepartmentPatrick K. Takahashi, Civil Engineering Department
University of HawaiiHonolulu, Hawaii 96822
. J
HAWAII GEOTHERMAL PROJECTWell Test and Reservoir Engineering
Progress Report for January 1977
TABLE OF CONTENTS
I. Wellhead Modifi·cations and January Discharge Test
II. Pressure Drawdown and Buildup Tests
1
A.
B.
C.
Pressu~e Drawdown Test . .
Pressure Buildup Analysis
Discussion ...
9
. 15
18
_J
Progress Report for January 1977HAWAII GEOTHERMAL PROJECT
Well Test and Reservoir Engineering
P.C. Yuen, B.H. Chen, D.H. Kihara, P.K. Takahashi
During January the modifications to the wellhead were completed andanother series of discharge tests were started. Pressure buildup and drawdown data from the November and December discharge tests were analyzed, andtentative conclusions reached on the state of the Pahoa Geothermal Field.
I. Wellhead Modifications and January Discharge TestBecause of numerous complaints about the noise from residents in the Puna
area and because of the need for additional safety conditions, several modifications were made to the wellhead equipment to improve its operation and to
alleviate the high noise levels produced. A specially-built muffler wasinstalled in place of the 24-inch horizontal discha;'ge line. The muffler, ofstandard design, is 6 feet long and made up of two annular regions, the innerone filled with cinders for absorption of noise while the outer region is empty.To reduce the noise generated by the low-frequency vibration of the separatorstacks, circular stiffeners were welded at two heights Qn each vertical stack.
The stilling basin for measuring the height of water flowing over the weirwas moved to a more convenient location for personnel monitoring. To facilitatethe insertion and removal of the temperature and pressure probes during dischargetests, a six-foot spool was added above the vertical valve. A steel platformwith stairway (which meets OSHA requirements) and a pulley and winch for manipulating the recovery tube have been installed in place of the temporary woodenplatform, which was deteriorating.
During the time that the wellhead equipment was being modified, the water
column level in the well was monitored with the results presented in Figure 1.On January 24, 1977, the 36th day after shut-in, the water level was at groundlevel. This curve is similar to that following the November discharge test.Two temperature profiles were taken during this buildup period and are shown inFigure 2. The temperatures in the wellbore have essentially returned to theirpre-flash values measured on December 8, 1976.
o
--~
....---
~/
/v
VII
a 5 10 15 20 25 30 351
4I
a
1000
<D><D 200...J
"'0c::::30l-(!)
~ 4000<Dco
IN -I <D
~6001
<D><D
...J
l-<D 800-03:
12/19/76 Days After Well Shut - in
FIG. 1. WATER LEVEL RECOVERY AFTER WELL SHUT-IN
1/24/n
Temperature In °c50 100 150 200 250 300 350
500
500
(/)'(1)CD
1000 ::.E.c.-a.(1)'
o
2000
600500400300200100
y\
\.
'~ -....~
....... ........~......
t\......~
",,~
" -' ..~........ ........,
If,1 -IT.-----. 1-3-77 ,+~
• a 1-15-77~I I
I I........
~~------ Bottom of well
I I I I-
I I ,
2000
1000
5000
6000
7000
10 3000if.5.c.-!4000
Temperature In of
FIG. 2. TEMPERATURE PROFILES AFTER DECEMBER DISCHARGE TEST
-3-
Equipment modifications were completed on January 25, 1977, and dischargetests began on January 26 following a warm-up period during which the well wasallowed to flow through a 2-inch bleed line. Initially, the well was allowedto surge for one-hour periods in an attempt to remedy suspected skin damage.During the three surges rock chips and drilling mud were ejected with the wellfluid.
Following this period of surging, the well was allowed to discharge withthe control valve wide open. A comparison of the characteristics of the flowsduring the early stages in the three discharge tests is shown in Table 1.There has been a steady increase in flow rate and wellhead pressure with eachsuccessive test.
Temperature and pressure profiles taken 48 hours after initiation of wideopen flow are presented in Figures 3 and 4. Within the accuracy of the instruments, these measurements indicate that the fluid in the wellbore is at saturationconditions throughout the wellbore so that the flow is a mixture of liquid andvapor. The slight change in slope of the pressure curve at 2090 ft. is due mostlikely to the change in the cross-sectional area of flow at the junction of the
slotted liner and the casing.Sound level measurements taken around the site show that the noise level
has been attenuated by roughly 7 dB--typically a reduction from 100 dBa to93 dBa inside the fenced area (50' x 80 1
) and from 87 to 80 dBa at the nearestpublic highway 120 1 away. In addition, the low frequencies associated withuncomfortable sensations of the chest and abdominal area (15 to 45 Hz) havebeen reduced. It appears that while some of the noise sources have been reduced,one important source, that of the circular stack's air column, has not and thatthis "organ pipe" remains as a primary source of sound.
A series of tests to determine well output parameters under throttled flowconditions was initiated by placing orifice plates of various sizes in the 8-inchportion of the discharge line. A 6-inch diameter orifice plate produced, asexpected, a rather insignificant change in the flow conditions. Four-inch and3-inch diameter plates did result in some change as shown in the data in Table 2.There is a substantial increase in wellhead pressure as the flow is throttled.A more complete table of data for a greater range of throttling will be obtainedin February and will provide preliminary design information required for selectinga wellhead turbine-generator.
-4-
TABLE 1
COMPARISON OF DISCHARGE TESTS AT 7 HOURS AND 25 HOURS AFTER INITIATION OF FLOW
AFTER 7 HOURS AFTER 25 HOURS
NOVEMBER DECEMBER JANUARY NOVEMBER DECEMBER JANUARY
II. Pressure Drawdown and Buildup TestsWhile data sufficient to assess a producible geothermal field can be
obtained only from a number of properly-spaced wells, some information can beobtained from a single geothermal well by utilizing the theory developed foroil and gas fields. A summary of the basic theory and references are given inHGP Engineering Technical Memorandum No.2, Geothermal Reservoir and Well TestAnalysis: A Literature Survey, 1974, by B. H. Chen.
During the two-week discharge test in November, data were collected whichpermit a pressure drawdown analysis, and after the one-week discharge test inDecember, data were collected for a pressure buildup test. Results from theanalyses of these two tests are given below.
A. Pressure Orawdown TestWellhead pressure vs. time plotted on log-log scales for type-curve
matching and on semi-log scales for a pressure drawdown analysis are shownin Figures 5 and 6, respectively. The initial pressure was obtained fromFigure 7. While these data can be used in a pressure drawdown analysis toobtain information about the geothermal reservoir, some skepticism must bedirected towards this analysis because of the following reasons:
1. The analysis is based on a constant production rate during thedischarge, and this condition was not held during the November test. Inorder to apply the theory, a normalized pressure was obtained by dividingthe pressure by the concomitant production rate.
2. There was some overpressure at the wellhead prior to the start ofthe test. Consequently, opening the valve took some effort and about 2 to3 minutes were needed to open the valve completely. Thus there is an uncertainty of that amount in the determination of zero time.
3. The theory is for bottomhole pressure whereas the data in Figures5 and 6 are for wellhead pressure. Thus the assumption must be made thatwellhead pressure is proportional to downhole pressure and the proportionalityfactor remains constant throughout the test.
With these restrictions and assumptions, several pieces of information canbe obtained. To normalize the pressure with respect to production the pressurerelation can be written as
FIG, 5, LOG~LoG PLOT OF NOVEMBER DISCHARGE TEST DATA
I I I I I I II I I I I I III I I I I I I II I I I I I I II I I I I I III
f- -
• ~
I- m=1.I1 PSi/klb/hr/CY~~..
-•
V•
I- -~
r- ~P, hr /q =5.23 psi/klb/hr~/ .. -
••
•I- -.-•
r- • -
-•'. •r- • -
-I I I I I I " I I I I I "I I I I I I I II I I I I " II I 1111111
1000100100.1 ITime t hr
FIG. 6. SEMI-LoG PLOT OF NOVEMBER DISCHARGE TEST DATA
2o
9
8'-
.J:;
.......
.0X....... 7(/)0-
CI)
'-:::)(/)
6(/)I Q)-" '--" 0-t
"00Q)
.c 5--Q)
::"0Q)
4N
0E'-0z
3
3530252015105
1\ P,:: 660 psig
\\\\1\
\•- •
•• •• •
•o
o
100
200
500
800
..600
700
~ 500If) .If)Q)
Q:
"t:J 400oQ)
.s::.-Q)
3=
-If)Q.
TIme t min
FIG. 7. LINEAR PLOT OF INITIAL DATA FOR NOVEMBER
DISCHARGE TEST
-12-
--------------
where Pi = initial pressure, psi
Pwf = flowing pressure, psi
q = production rate, std bbl/day~ = viscosity, cpB = formation volume factor, res vol/std volk = permeability, mdh = formation thickness, feett = time, hr$ = fractional porosity
Ct = total system effective isothermal compressibility, psi- l
rw = well radius, ft
s = skin effect factor
The left side of equation (l) is a linear function of 10910t so thatP. P f1 - Wa plot of q vs. 10910t will yield a straight line with a slope, m,
psi/bbl/day/cycle, where
and this equation can be used to calculate the permeability-thickness, kh.Equation (l) can also be used to calculate the skin effect factor, s.
Letting Plhr be the value of Pwf for t=l hour on the correct semi-log straightline, equation (1) can be rearranged to yield
(3)
By using (3), the pressure drop due to the skin effect can be calculated from
~Pskin = 0.87 Imlsq
-13-
(4)
and the flow efficiency
P - P -!:lpi wf skinFE = -..1.9----
Pi - Pwfq
(5)
With the assumptions made previously, a log-log type-curve plot ofpwf vs. t for the November test is shown in Figure 5. The two unit-slope
Pi -q
lines shown verify the existence of we11bore storage effects. From the end ofthe second straight line, it appears that the semi-log straight line or theradial flow period started at about 10 hours after the test was begun.
Pi - PwfFigure 6 is a semi-log graph of q vs. 10g10t. An analysis of the
plotted data shows that the permeability thicknesskh = (162.6) (24 hr/day) (0.09 cp) (1.5 res bb1/std bb1)
(350 1b/bb1) (1.11 x 10-3pSi/1b/hr/cyc1e)
kh = 1356 md-ft
and if the thickness of the producing layer is as~umed to be h = 1000 ft.,then the penneabi1ity
k = 1.4 md .
The skin effect factor
s = 1.15 [5.23X10-3
- 10g10 1.4 2 + 3.23J = -0.861.11x10-3 6 8 755
(0.03)(0.09)(8x10- )( 24 )
The small negative skin effect factor suggests that skin damage is not present.Therefore, the flow efficiency of the well is approximately 1, or the well isdischarging as much as it is able to produce.
The minimum drainage area for the duration of the November flow test can beestimated to be
A = 0.000264 (1.4)(3.36) = 1.15 x 108 ft2(0.03)(O.09)(8x10-6)(0.05)
Thus the minimum volume reached during this discharge test was
B. Pressure Buildup AnalysisAs with the pressure drawdown test, the pressure buildup test employs the
standard methods used in petroleum and gas field analysis. The end of theDecember discharge test permitted a pressure buildup test. Bottom-hole pressureswere taken by two Kuster KPG pressure elements and recorders in tandem to ensurethat pressure data were acquired since considerable difficulty had been experienced with equipment malfunction because of the very high temperature.
Figure 8 is a log-log type-curve plot of {Pws - Pwf ) vs. t. It shows twodistinct we11bore storage effects as in the pressure drawdown test; the topof the second we11bore storage effect is indicated by the arrow A. The rule ofthumb used is that the onset of the radial flow period on the conventional semilog straight line is 1 1/2 log cycle beyond A, which is indicated by the arrowB. This time is approximately 70 hours after well shut-in. Figure 9 is a semi-
C. DiscussionTable 3 summarizes the preceding analyses of the pressure drawdown and
buildup tests. The permeability thickness figures from both analyses aresimilar, but the skin effects and flow efficiencies are widely divergent. Theassumptions for a pressure drawdown analysis include the production of fluidat a constant rate, which is difficult to satisfy in practice. In order toapply the theory, the pressure data were normalized by dividing by the production rate, which can be questioned for its validity. On the other hand,the pressure buildup analysis has no similar, difficult assumption to satisfyin practice. Thus more reliable conclusions can be drawn from the pressurebuildup test and analysis.
In a very preliminary way the pressure buildup test indicates that thereservoir is tight (low permeability of perhaps less than 1 millidarcy) and thatthe well suffers from significant skin damage, resulting in a discharge rate of
only 65% of what it is capable. This latter tentative conclusion is supportedby the data in Table 1, which shows that the flow rates have increased with eachsucceeding test. This may have been a result of the initial surges in eachtest, which either removed the baked-in mud and thus reduced the skin damage,or possibly induced stress-caused microfractures.
-18-
TABLE 3COMPARISON OF PRESSURE BUILDUP AND DRAWDOWN TESTS